The Fault Diagnosis Method Based On Multivariate Statistical Analysis And Its Application

With the increasing requirement on safety, reliability and effectiveness of automation devices, study on the problem of fault diagnosis has received great attention and been one of the most active research topics during the past three decades. Multivariate statistical analysis has been considered as an important method of fault diagnosis due to its independence of process model and the availability of mass data in industrial process.The main purpose of this thesis is to make further study on multivariate statistical analysis based fault diagnosis method and develop its application to the on-line fault detection and identification of a Three-tank system. The main context consists of the following three parts.By using the fault information in principal component subspace (PCS) instead of residual subspace (RS), the T~2 statistic based fault reconstruction technology is used to improve the ability of fault identification and, based on this, the fault identification is implemented via fault reconfiguration.Based on the statistic index used in fault reconstruction, we briefly classify the fault reconstruction based methods into two kinds, named the SPE-based and the T~2-based methods. With suitably selecting the fault isolatable magnitude in PCS an RS, the fault isolation ability of these two methods will be compared with each other and, by a combination of them, the accuracy of fault identification can be increased evidently.A system structure of fault diagnosis by using PCA is proposed to improve process monitoring performance and, based on this, the on-line process monitoring of a three-tank system based is developed, which can be used to realize model parameter adjusted on-line, fault detection, fault alarm and fault identification.As a summary, this thesis is to make further research on fault identification by...